There's a tendency of smart people to try to overfit technology to a situation. Sometimes a technology should impose structure of its own. And sometimes it needs to intentionally not rigorously address a use case or it becomes inflexible and people just work around it
One of the hardest product lessons for me to remember to this day is that just because something is a good solution to a problem doesn't mean it's a good product. And vice versa.
Not sure if it's a product yet or just a toolkit for myself when I need to address niche needs, but I'm getting close to asking around for others who may have a need and some interest in trying it
Recognizing I need to talk about projects as they take shape: I'm working on a flexible way to build one-off workflows that *also need to be easily configurable by a user*. Think geographic localization that users need to be able to control intuitively with a map
Open source software never ceases to feel like a super power. You mean someone else had this same need, solved it, and just… let’s anyone use and adapt it?
I'm happy to announce that as a counterpoint to "build in public" I've decided to build in secret. I will stock up on supplies, retreat to a cave, and only emerge when the product is built and winter is over... Oh. Hibernate - what I want to do is hibernate.
I've been meaning to find this quote I think about in relation our information ecosystem:
""Where does a wise man hide a leaf? In the forest. But what does he do if there is no forest? ... He grows a forest to hide it in ... A fearful sin" - G. K. Chesterton
I wish there were better real world answers to open source sustainability. As much as I believe in the benefits and don’t like being a gatekeeper, I’ve often chosen not to open source things I’ve built that I use in my business… but I don’t always feel great about that decision
I often build small, niche APIs (usually 1-2 “glue” flows). So with all the advice about premature optimization in mind, I built a toolkit to cover stuff I do over and over. Still deciding whether it’s an internal tool or something to share. Obligatory: https://t.co/ocZRDhTC0a
What’s the great AI work being done in narrowly constrained domains? I get that “code anything!!” is a huge market, but I’m really interested in using LLMs for more accessible interfaces on things like build processes that are testable and also represent a barrier to entry
It’s been too long since I used a screen reader for an afternoon. Best practices are a convenient shorthand, but there’s nothing like actually trying to get things done in a modality you’re not used to
One of my first memories of web development was delight that I could just save a few lines out of a text editor and have a working page. I fully appreciate modern web build pipelines, but I wonder what we lose when we abstract away some of that “tactile” directness
As a parent I’ve learned that attention == good, and “negative” attention, while sometimes necessary, isn’t the consequence I think it is. I wonder if there are any parallels there with journalism and communication….
I think there’s scope for some AI in analytics, but let’s not forgot that the most important step is *what questions you ask.* And that fundamentally can’t be automated, not because our tech isn’t smart enough but because it ends up not being about information, but what we value
If you want to get ahead, sometimes you have to grind and just *make* time for progress. Work late, get up early. Does it feel good? No. Is it compatible with maintaining your health? No. But do you get things done in that time? Also no.
I guess I’m saying take care of yourself
When I hear praise for uncompromising high standards, not only do I make a judgement about life circumstances that enable that outlook, but I also wonder about lack of imagination of how things could be even better. Everything is a compromise (done is better than perfect, etc.)
I worry clients will wonder why I'm posting to social when I owe an email or a platform feature, so let me assure you: posts made from 100% recycled, otherwise unproductive time: most recently, lying behind the couch wondering if my son has forgotten we're playing hide and seek
Being early doesn't seem to me to correlate that strongly with success. Definitely not compared with, say, deep audience understanding and connection. Sometimes being first into the surf gets you smashed on the rocks and you're better off waiting to understand how boats work
Let's say AI is the Next Big Thing for a field like journalism, on par with the internet (not saying it is... we've had some, uh, false starts -- blockchain anyone?). Does it necessarily follow that we should adopt it ASAP?